1. Field of the Invention
The present invention relates to technology for arranging storage regions of storage logically into tiers, distributing file groups through relocation into each tier, and managing the relocation of files belonging to each group into collections of files belonging to any tier as file groups.
2. Description of Related Art
A storage system is equipped, for example, with at least one storage apparatus referred to as a disc array sub-system, etc. In this type of storage apparatus, disc drives such as, for example, hard disc drives or semiconductor memory drives etc. are relocated in an array so as to enable storage regions to be provided based on RAID (Redundant Array of Independent Inexpensive Disks). A host computer then accesses logical storage regions supplied by the storage apparatus and carries out reading and writing of data.
In the related art, a storage system provided with a plurality of storage apparatus having one or more volumes, a virtualizing section for collectively managing volumes in the possession of each storage apparatus in a virtual manner, a storage section for storing volume attribute information for managing attribute information of each volume, and a relocation section for relocating a specified transfer source volume to a specified storage tier between a plurality of storage tiers generated based on a plurality of policies set in advance and volume attribute information is proposed (refer to patent document 1).
[Patent Document 1]
Japanese Patent Laid-open Publication No. 2006-99748.
With storage environments that are structured in tiers of storage systems of the related art, it is necessary to relocate files to each storage tier in an optimal manner taking into consideration utilization frequency and importance of the data.
Namely, storage systems of the related art adopt a configuration taking access frequency of file group units where a file or plurality of files are collected together as a reference, and relocate files to any one of the tiers every unit. This is therefore not sufficient for relocating files to file groups in an optimum manner.
Specifically, the file group is utilized at irregular intervals, and utilization every one time is for a comparatively short time. However, a state is adopted in the storage system where files within the group are used overall, and if the deciding of the file used first in a one time utilization of the file group is not taken into consideration, the following problem occurs.
For example, in order to promote files belonging to a file group to the uppermost tier, an access of a certain extent is necessary, and a time lag occurs until use at the uppermost tier is possible. Because of this, when a file of the file group is promoted to the uppermost tier, it is common for the operation using this file group to have advanced to a certain extent. However, when the access frequency falls, it is possible for files belonging to this file group to be demoted to a lower tier. Further, it is also possible for the group as a whole to be relocated through interchange at the uppermost tier or a lower tier. There are therefore cases where a comparatively large reduction or increase in capacity occurs with regards to the storage region between the transfer source and transfer destination at the time of file transfer.
It is therefore the object of the present invention to collect files belonging to a file group at the uppermost tier when a specific action occurs with regards to a file group. Moreover, it is a further object of the present invention to always maintain optimum file promotion conditions or file demotion conditions with regards to the file group by dynamically changing the files subject to observation (key files) within the file group.